Hybrid vehicle smog failure rates (database)
 

A searchable database of 2015 California Smog tests (11 million of them) was posted by the newspaper the Sacramento Bee last April.

Last year was the first year hybrid cars needed to be tested. Some people anticipated a rude awakening for hybrid owners, assuming hybrids cars would often fail because they had never been tested.

The database shows failure rates for vehicle models, year-by-year, and compares them to statewide average rates of failure for the same model year.

I searched for several Honda and Toyota hybrid models and found the Toyotas seem to do better. Honda hybrids failed more often than average cars, while Toyotas almost always failed less often. However, there was no great wave of failure, like sensationalists predicted.

For example, while 3.5% of all 2009 cars failed in 2015, 9.1% of the 396 Civic Hybrids tested failed. Civic Hybrid models from 2003-2013 failed much more often than other cars from those same model years.

The Insight and CRZ data sets were also relatively small. But the first gen Insight (2000-2006) sometimes had higher than average and sometimes lower than average failure rates in 2015. The same was true of the 2010-2014 Insights and the Honda CRZ.

I wondered if the hybrid testing procedure is more strict, producing more failure. But the Prius generally failed much LESS often than the statewide average of all cars. For example of the 16807 Priuses from 2005 tested in 2015 only 2.8% failed, while 6.7% of all 2005 vehicles failed. And the Lexus CT200h, a car with Prius guts, also generally had lower failure rates than the statewide 2015 average.

So the upshot seems to be that the Prius/Lexus hybrid emission systems are probably more reliable, long term, than the Insight/CRZ systems as well as all standard ICE systems California tested in 2015. It does not mean the Hondas pollute more because failure rates don't tell us why cars failed or how much pollution the cars that passed were actually emitting.

Check out the SacBee database for yourself:
How often does your type of vehicle fail smog tests? See answer here | The Sacramento Bee

This one line makes this entire story a non-story.

I would be fascinated to know exactly why the vehicles fail because it speaks specifically to a vehicle I drive (even though we don't have emissions testing where my car is registered, so it's even more a non-story as far as I'm concerned). Is there another database that calls out that information?

Well I don't think that's correct elhigh. Inconclusive as to causes, sure. And I pointed that out myself. But aggregate, general data is good for identifying trends and promoting further investigation. As this simple data does. I am sure CARB has the data you describe because all emissions reports are immeditately sent to them electronically. But this data helps draw attention to the trend. And we can theorize further questions. Since the Insight and the Prius compete against one another for the same segment of the car buying public it would be hard to explain the opposite outcomes in longterm emissions reliability through driver factors alone. Chances are there is something about their systems too. We already know that Honda and Toyota are using quite different hybrid technologies. So the data seems to show that there is something worth further investigation.

This all seems rather puzzling, as the hybrid system has nothing whatsoever to do with emissions - at least in any way that could be tested without a dynamometer or road test.

Don't know what California's testing involves*, or why they'd only start testing hybrids this year, since they've been around since 2000, but the only thing I can think of that seems at all realistic is an automatic failure if the OBDII system has thrown a code that turns on the Check Engine light. On my Insight (2000 model), that happens fairly often - say once a month or so on average - for reasons involving the battery and/or DC-DC converter. Nothing to do with the IC engine or its emissions, IOW.

*Nevada's smog test for my Insight is to simply plug a scanner into the OBDII port.

I agree it is odd. And I don't know a lot about it, except this. The system in place for testing hybrids is different than the system for testing ICE vehicles like mine. It does not have a sniffer in then exhaust. That's because, I have read, many hybrids will not idle long enough for the test. So, I actually think the hybrid systems could still turn out to play a role (not THE role). Maybe the Hondas damage the catalysts more quickly over time? Honestly I don't know, but the pattern seems unmistakable: honda hybrids are not passing as often as toyota hybrids. And the Honda Hybrids do worse than other non-hybrid Hondas. The 2005 and 2006 accord hybrids are the worst of the the accords in those years. Meanwhile the 2007 and 2008 Camry Hybrids do about as well or better than non hybrid Camrys. Just saying... that's a pattern.

Not puzzling to me - hybrids lower emissions significantly as they optimise the running of the ICE, both how long and at what load it runs. The Toyota system augments the ICE significantly in low load / idle scenarios and plays a huge part in the overall emissions. One example is where the ICE warms up in "low emissions" mode while the electric motor / HV battery propels the vehicle.

From a reliability standpoint, these observations are no surprise to me - how many articles have I read on this forum and others about faults in the Honda IMA hybrid systems (mainly batteries), whereas the Toyotas rarely have issues. Also, how many Prius engines have needed EGR cleanouts to make them run properly?

I know that I may seem biased as I own 2 hybrid Lexus vehicles, but the technology seems vastly superior to the Honda IMA.

Simon

But there's no way to check most of that except with a road/dynamometer test, and to check what happens during warm-up, you'd have to bring the car to the check station and leave it for hours to cool down before running the test. And that's Toyota: Honda's IMA doesn't even kick in unless the engine is under load.


I've always thought that the Toyota system was far inferior, at least as far as driveability. But that's really sidestepping the question of whether these are actually real emissions failures, or bogus ones due to Hondas being prone to throwing diagnostic codes.

Absolutely James... loke what elhigh and I were getting at... this data does not tell us the cause of the failure... and the systems trouble could be something relatively harmless, like DTCs when the test is being done or something about how the test protocol and the system's "fit" with it. Would be cool to have more data.

Q&As: Smog Check Program

According to this page, a CEL will fail you. I wonder if Hondas don't lean out when revving, possibly increasing NOx, given that they have lean burn. I also wonder if different vehicles don't have different thresholds - the CVT Insight has SULEV certification, while the 5MT has ULEV.

In Texas, the emissions test they did consisted of plugging the computer into the OBD-II port (at least for vehicles that have one) to check for CEL/codes and checking that the gas cap held pressure. If California is checking for CEL or codes, the higher rate of battery failure in Honda hybrids probably accounts for most of their failures. Also, a Honda hybrid can still drive with a dead battery, while a Toyota hybrid couldn't even start up to drive to the testing station. Only a new battery would allow it to be driveable again, which would clear the code.

And could California add a safety inspection? I saw many vehicles with burned out lights when I lived there. I would rather the cars around me be checked for functioning brakes and lights than pollution.